276 ALASKAN GLACIER STUDIES
narrow gully, deepest at the arch of the uplift, where there was a little cascade. The whole arch is 100 feet or so in width.
That icebergs were formerly discharged in this cove, when the glacier extended out farther than now. and the water was not yet so shallowed by deposition, is indicated by an abandonee! wave-cut cliff on the Heather Island side of the lagoon. At the top of this cliff were trees fifty to seventy-five years old; but the face of the cliff is partly covered with a growth of peat, growing in which and at the base of the cliff were trees fully twenty-five years old. The cliff-cutting had, therefore, ceased over twenty-five years before 1909.
It is not known exactly what changes in detail took place in this part of the terminus of Columbia Glacier between 1899 and 1908, but between 1908 and late August, 1909, there was sufficient advance of the terminus so that the details shown in Grant's 1908 photographs were entirely destroyed by August, 1909. Marked thickening of the ice in this part of the terminus is demonstrated by a comparison of the Gilbert photograph from the islet north of Heather Island in 1899, with Grant's 1905 and 1908 photographs from the same site. These show that there was not only an advance and thickening between 1899 and 1905 but that the advance ceased, the ice thinned again by ablation and was again thickened by advance between 1905 and 1908, the thickening continuing very markedly in 1909 and 1910.
In Gilbert's 1899 photograph (PL CVI, A) the part of the ice front beyond the eastern timber belt shows as a low, irregular surface, completely mantled by debris. In Grant's 1905 photograph (PL CVn, A) this area is occupied by crevassed ice, at least fifty feet thicker than in 1899, judging by the height to which it rises upon the nearer inclined tree trunks through which it shows. It may be even thicker. Following this advance and thickening there had been some ablation, for the terminus was again partly mantled with debris and the crevasses partly healed. This advance had, therefore, ceased before 1905 and had not extended out any farther than the 1892 maximum, as shown by stones in the foreground of the 1899 and 1905 pictures and by Grant's measurements of the retreat of the ice front several hundred yards to the west. Comparison of Grant's 1905 and 1908 photographs show that the retreat continued some time after 1905, for the surface seen through the trees was lowered very considerably by ablation between 1905 and 1908, though still thicker than in 1899. The surface to the left of the trees in the 1908 photograph (PL CVH, B), shows a great crevassed dome of clear ice, indicating that a new advance was then in progress, this being also shown by the advance of the glacier margin as measured by Grant. The thickening between 1905 and 1908 cannot be determined closely, for we do not know how much lowering by ablation there was after 1905 and before the 1908 advance began; but it was evidently not less than fifty or seventy-five feet, and may have been much more. This 1908 advance continued into 1909 and 1910, being observed by Grant in June (PL CVIII, A) and by the National Geographic Society's expedition in August, 1909, and on our two visits in 1910.
Thickening and forward movement of this part of the terminus is also shown by comparing Gilbert's 1899 photograph from the 1497 foot station (I) on the hillside to the east with our own taken in August, 1909. Projecting the terminus in the two photographs forward against the islands north of Heather Island it is evident from Gilbert's map that the net advance in the last ten years is approximately five-sixteenths of a mile or about 1600 feet. It is not known how much of this was between 1899 and 1905